Forum adverts like this one are shown to any user who is not logged in. Join us by filling out a tiny 3 field form and you will get your own, free, dakka user account which gives a good range of benefits to you: No adverts like this in the forums anymore. Times and dates in your local timezone. Full tracking of what you have read so you can skip to your first unread post, easily see what has changed since you last logged in, and easily see what is new at a glance. Email notifications for threads you want to watch closely. Being a part of the oldest wargaming community on the net. If you are already a member then feel free to login now.

Yes, they can sense, but they can't see can they. Not like a human eye can see, process and understand. They can only sense what they are programmed to sense. They are incapable of understanding. 'Bizarre faith in the superiority of humans'
I think that might be well placed. I'm going to take that over blindly surrendering everything to machines, created by humans in the first place.
You're not going to convince me that computers are superior to humans.
They may be able to beat chess grandmaster, but that's not the real world.



Automatically Appended Next Post:
And yes I'm writing this on a computer, but I don't think even you can deny that if I said the same words to you face to face, the communication would be superior.

 queen_annes_revenge wrote:
Yes, they can sense, but they can't see can they. Not like a human eye can see, process and understand. They can only sense what they are programmed to sense.

And? They can be programmed to recognize what they need to recognize. They don't need to "understand" it in some philosophical way, they just need to process the data correctly and take the appropriate action. Current data processing systems need work to get an acceptable rate of hazard recognition, but the automated system has significant inherent advantages there. It doesn't look down at its phone and take its eyes off the road, it isn't limited to looking in a single direction, its view isn't obstructed by the body of the vehicle, etc. Once image recognition technology progresses a bit more the automated system is going to be far, far better than any human driver at the task of "detect a pedestrian hazard in time to avoid a collision and take appropriate action".

Also, you're going to have a hard time convincing anyone of the value of human eyes when there's a long list of accidents where the driver "just didn't see them" and killed or injured someone. You're defending a badly flawed system.

I'm going to take that over blindly surrendering everything to machines, created by humans in the first place.
You're not going to convince me that computers are superior to humans.

Like I said, blind faith in human ego. You have no evidence, just a stubborn insistence that humans are magically better because you want it to be true.

And nobody is asking you to blindly surrender everything to machines, that's what testing is for. Automated vehicles will undergo extensive testing to prove that they can be at least as safe (on a per-passenger-mile basis) as human drivers before they are approved for full-scale use. But once they have passed this test you are no longer justified in objecting to them replacing human drivers. Continued defiance out of blind faith and ego means that the blood of every additional dead victim is on your hands.

And yes I'm writing this on a computer, but I don't think even you can deny that if I said the same words to you face to face, the communication would be superior.

You're missing the point. The fact that you are successfully writing this instead of staring at a lump of dead silicon is a demonstration of a lot of machines working near-flawlessly, to the point that any error is an unexpected outrage instead of the default state of things.

Queen its true that computers still need to interpret what their sensors see of the world and that there is still some further study required there and that its not perfect.

Thing is humans aren't perfect either! Consider all the people who have seen ghosts, illusions, optical oddities and such - all without the aid of drugs or medical conditions.

The human brain can be tricked, heck a vast amount of magic tricks are all about tricking the brain into seeing and not seeing things.



You also state that a computer reacts based upon volume of data and that is true, but that's the very same way a human reacts. A human with more training and more experience has more incidents to draw conclusions for. You as an ambulance driver are ahead of the curve in having both additional training and additional driving at high speeds through more complicated driving situations. So you can readily process quite a lot more possibilities because you've got a higher range of experiences to draw from.
Joe in the regular car next to you doesn't have those additional layers and Jane passed her driving test after 3 weeks of intensive training and has only been on the road for 4 weeks through summer and has never seen proper night, rain, snow, ice driving.

Yet you'll all be on the road at the same time, even with nothing going wrong its possible for one of you to make a wrong judgement call at a turning or intersection. Jane might well make a wrong call on a car clearly moving to turn, but not using its signal. You might not be stumped by that at all, you've learned "card body and position" language to a higher level.




A computer in a car has the bonus that once its learned its learned and every other single car with that same computer also learns that experience. It won't forget because it hasn't driven in night for 6 months; it won't have to go through a near crash because car 101 already did it and has passed that experience to the rest of the fleet.

Basically the technology rests upon thresholds of understanding for the machine, once it can more readily and accurately identify a person and another car it becomes more trivial to give it wider ranging sensors that look further and further ahead. Plus if every other car is automated there's nothing to stop them speaking directly to each other. A sudden braking a mile away due to a problem can be passed right back along both sides of the road; all cars are then slowing down long before any of them can see the incident (human or machine). That's a level of communication and safety that can likely only be matched right now with lorry drivers and their radios (even then I don't know if they can all easily talk with each other from different companies and such)



This isn't computer worship, its accepting that a machine can achieve something better and safer than a human can. OR at least that the machine has the potential to do so with sufficient advance of the technology.

Not from you no, but peregrine seems well and truly in the grip of the omnissiah.
And like I keep saying, it's just a case of this or that is still just an idea. The robot I use is programmed not to conflict with its own frame yet that constantly fails, and thats made by one of the world's most successful aerospace companies. So saying all cars will talk to each other, well that working successfully seems like a distant pipe dream. I'd be interested in knowing some of the experience that the people blindly advocating complete autonomy have with actually working with automation and why this makes them believe what they do. Because if the only reason is that computers are pretty good at doing things, that not really the case when it comes to the world away from a desk.

I think Peri is just looking (at least*) 10 years in the future whilst you're looking at the now/past a bit. Which is part why your viewpoints are so varied.

He's thinking of the future when the machine is ready; whilst you're looking at today when its clearly not.




*and likely more honestly. I'd wager we'll see automated trains before cars; at least if the companies can get past the driver unions and such.

My viewpoints aren't varied, I'm just trying to cover all the questions posed by this issue.
And that in itself is another question.. What do we do with all those people who depend on driving for employment?

And I pray that in 10 years time this will have been seen as a terrible idea and canned, with the research capabilities going to more productive endeavours. I can only hope.

Well depends. I can see in services like train driving and buses that they'd likely retain the driver to oversee things and be there for public relations. They'd likely shift to being more a porter in some ways and a guard than the driver.

Meanwhile for things like lorries, again, I could well see them keeping staff on as drivers to oversee things. Plus a lot of lorries drive in country roads and deliver to out of the way places. So I could see drives being kept on for road networks that might be off-grid as such for self driving or for parking and manoeuvring. Plus they are there as a human element to oversee deliveries are made and cargo is kept in check.

So for all those I can see the role of the driver changing, but the actual job remaining part of the setup.


Taxies might suffer the most since that is purely driving and once you've got A to B programmed in the taxi should get you there on its own. However there's a good number of alternative destinations that might not work on a sat nav. Or the person might want to go to a location not an address. Again we might see the driver remain but the role shift and change.

So yes there might well be job losses, same as there is with many forms of automation. The messy period is where we transfer from a very heavy human working population to a very heavy mechanised.


Of course as I cited earlier somewhere else, some factories are putting humans back and taking machines out because changing designs significantly requires rebuilding most of the factory. That's fine when you first built it, but its a huge cost when you've already invested in it all once over. However in part that's showing that many robotic assembly machines are both very specific and limited in adaptability and also require human operators to make significant design changes.

 queen_annes_revenge wrote:
And that in itself is another question.. What do we do with all those people who depend on driving for employment?

What happens with any field that becomes obsolete as technology advances? Seems like a good argument for socialism, but solutions to unemployment are rather off-topic from here.

And I pray that in 10 years time this will have been seen as a terrible idea and canned, with the research capabilities going to more productive endeavours. I can only hope.

Their blood is on your hands.

 Grey Templar wrote:

Indeed. Self-driving tech is going to be sacrificed on the alter of litigation. Someone/everyone will get sued and the different corporations will realize that self-driving vehicles are a terrible idea from that standpoint, so they'll ditch the idea like a hot potato. It'll be just another fad that comes and goes in the mid/late 21st century.

The amount of money corporations will make from simply firing 99% of their drivers for the next thousand years will easily cover any amount of money lost to litigation.

These lines of thinking about insurance are incredibly dumb. In my city alone something like three million people A DAY ride public transportation via buses and trains. If someone within that three million gets hurt, are the passengers held liable? Oh gak, no they're not. What about boat cruises? What about planes? Hell, what about cars? If the brakes in my car suddenly stop working and it comes out later that a defect in the car's design was responsible for my brakes seizing, am I held liable for a resulting accident, or will the company that made my car be held liable for selling me defective equipment? Did Toyota (or maybe it was Honda, w/e) say "whelp I guess we're not making cars anymore lol" after that mass recall they had to do around a decade ago due to thousands of their cars having a fatal safety defect? No, of course not. Car companies will never give up billions of dollars in profit because they lost at absolute most a hundred million in litigations, and every public transportation system in the world has dedicated insurance companies with clearly defined statutes on who's liable when a train runs someone over or a plane's autopilot fails.

You people are asking questions about a topic that's existed for at least 40 years.

- - -

Anyway, anyone who is against automated driving is a psychopath and I hope you will never have to experience losing a loved one to a DUI driver or a boomer who fell asleep at the wheel.

 Peregrine wrote:

 queen_annes_revenge wrote:
And that in itself is another question.. What do we do with all those people who depend on driving for employment?

What happens with any field that becomes obsolete as technology advances? Seems like a good argument for socialism, but solutions to unemployment are rather off-topic from here.

And I pray that in 10 years time this will have been seen as a terrible idea and canned, with the research capabilities going to more productive endeavours. I can only hope.

Their blood is on your hands.

Yawn.


Automatically Appended Next Post:

 BlaxicanX wrote:


Anyway, anyone who is against automated driving is a psychopath and I hope you will never have to experience losing a loved one to a DUI driver or a boomer who fell asleep at the wheel.

What a ridiculous statement. You and peregrine must've been sharing straws.

What happened to the people that raised all the horses or made carriages?

How many people die because of Truck drivers falling asleep at the wheel, or accidents from someone in a hurry making a dumb decision?

Hell, Canada had a pretty good example of what happens with the Humboldt Broncos incident. Saying that Humans will react better than computers seems kind of silly when we see plenty of moronic deaths on a daily basis in traffic.

Of course, your entire argument seems to boil down to "Computers bad" and "You aren't taking my driving away!".

If you want to straw man it, then yeah I guess, but then my respone would be, 'humans are unsafe, computers amazing, if you drive you're a murderer'

What I keep seeing here queen, is exactly what people are saying: you are taking a very luddite position that kinda boils down a strawman in itself.

What many/most of the other users are saying here is, this is not a currently viable technology, however work is being done to make it viable, and when that happens, we will be ABLE to do away with incidents like the Humboldt Broncos and thousands of other non-related DUI incidents. Feth, kids will be able to get on their school bus safely, because in a hurry donkey caves won't be flying past the flashing red lights. That isn't now. . . that is the future.

I've stayed multiple times in this thread. That my rational position is that there are too many variables, unanswered questions, and flaws in the technology for us to be blindly accepting of it at this point in time. (independent, though related to my personal opinion of not liking the idea of autonomous vehicles)
Other points I've tried to raise are just related concerns.

 queen_annes_revenge wrote:
I've stayed multiple times in this thread. That my rational position is that there are too many variables, unanswered questions, and flaws in the technology for us to be blindly accepting of it at this point in time. (independent, though related to my personal opinion of not liking the idea of autonomous vehicles)
Other points I've tried to raise are just related concerns.

Nobody is arguing for blind acceptance, that's what testing and study of safety records is for. Automated cars are not going to be in full-scale use until they have passed the testing stage. But we should expect them to pass eventually, and your opposition to life-saving technology out of blind faith and ego is morally unacceptable.

Lets not forget how fast technology is advancing these days!

Go back 20 years ago and the idea of building a robot that could walk and move on its own was pretty high end and dreamy future stuff. Now Boston Dynamics have a free standing robot that can walk, run, jump and even jump up obstacles on one leg!

So yes self driving cars can't do it now, they can't do it today; but they are getting close and in 10 or 20 years the technology on all fronts will have moved on a lot. Heck if they can ever crack functional affordable and reliable quantum computers that would lead to a vast revolution in computing power which on its own could give machines an infinity more powerful ability to process vast volumes of data (eg visual data for a self driving car)

 Peregrine wrote:

 queen_annes_revenge wrote:
I've stayed multiple times in this thread. That my rational position is that there are too many variables, unanswered questions, and flaws in the technology for us to be blindly accepting of it at this point in time. (independent, though related to my personal opinion of not liking the idea of autonomous vehicles)
Other points I've tried to raise are just related concerns.

Nobody is arguing for blind acceptance, that's what testing and study of safety records is for. Automated cars are not going to be in full-scale use until they have passed the testing stage. But we should expect them to pass eventually, and your opposition to life-saving technology out of blind faith and ego is morally unacceptable.

You have no right to label me morally unacceptable when your only argument is that because I don't accept that computers are safer than humans, that somehow makes me a murderer. I think that straw man has been well and truly established.

Peregrine wrote:What happens with any field that becomes obsolete as technology advances? Seems like a good argument for socialism, but solutions to unemployment are rather off-topic from here.]And I pray that in 10 years time this will have been seen as a terrible idea and canned, with the research capabilities going to more productive endeavours. I can only hope.

Except that without employed people the companies will have nobody to purchase their robot made product, which will dry up resource supply AND that batch of other people's money that your socialist views depend on so desperately. No CEO is going to wake up and decide to give his money away for the greater good. If it came down to it, they'd take the entirety of their fortune and retire in Bora Bora, staring at caramel colored tiddies while all the socialists are forced to beat each other over the head for basic resources. We have historical examples of this happening. This isn't deluded fantasy, it's human condition.

Peregrine wrote:Their blood is on your hands.

BlaxicanX wrote:Anyway, anyone who is against automated driving is a psychopath and I hope you will never have to experience losing a loved one to a DUI driver or a boomer who fell asleep at the wheel.

Hyperbole and sensationalism NEVER sell your point unless its a prog argument like this. Had someone tried to sensationalize that... controversial reproductive rights issue... you two would be the FIRST to lash out at them for sensationalizing.



And to address BlaxicanX directly: I've seen FAR more accidents, including fatal accidents, caused by empty headed millennials who can't keep their nose out of Facebook on their phone in traffic than I've seen by geriatric narcoleptic "boomers". Your bigotry is ill placed, and makes you look like a complete idiot.

 Just Tony wrote:
Except that without employed people the companies will have nobody to purchase their robot made product, which will dry up resource supply AND that batch of other people's money that your socialist views depend on so desperately. No CEO is going to wake up and decide to give his money away for the greater good. If it came down to it, they'd take the entirety of their fortune and retire in Bora Bora, staring at caramel colored tiddies while all the socialists are forced to beat each other over the head for basic resources. We have historical examples of this happening. This isn't deluded fantasy, it's human condition.

I don't think you understand how the economy works. Capitalism requires customers, but every business is out for its own interests not the collective good of the economy. As we have seen over and over again, if a company sees a way they can take advantage of automation to cut a bunch of employees they won't hesitate to do so. Every company will be saying "someone will keep people employed, I don't need to worry about it" even as the overall system collapses, because failure to do so means dying even faster than the average as the competition (which is cutting useless workers and embracing automation) undercuts their prices and takes all of their sales. It's absurd to suggest that companies are going to voluntarily pay more to keep redundant employees around so that they can maybe get some of that salary back in the form of sales, or that the same CEOs who will cut and run at the first sign of socialism taking their wealth will voluntarily pay those extra salaries out of pure altruism towards the good of the economy.

And that whole "take the money and run" plan depends on the state allowing wealth to leave, and the world not collectively saying " those guys" and erasing Bora Bora from the map now that all of the troublemaker CEOs are collected in one neat easily-bombed package. After all, why beat each other over the head for basic resources while the people who caused the problem live in luxury when you can nationalize all of their wealth and beat them over the head?

Hyperbole and sensationalism NEVER sell your point unless its a prog argument like this. Had someone tried to sensationalize that... controversial reproductive rights issue... you two would be the FIRST to lash out at them for sensationalizing.

No, I'd be the first to lash out at them for being factually wrong. I don't really care about sensationalizing when there are plenty of better reasons to attack someone.


Automatically Appended Next Post:

 queen_annes_revenge wrote:
You have no right to label me morally unacceptable when your only argument is that because I don't accept that computers are safer than humans, that somehow makes me a murderer. I think that straw man has been well and truly established.

Watch me. Your beliefs are morally unacceptable.

It doesn't make you a murderer, but it makes you responsible for those additional deaths. Your anti-computer beliefs are not based on evidence, they're based on emotion and ego. And you are opposed to life-saving technology because of that emotional response and inability to accept facts. Every day that people like you delay the adoption of life-saving technology beyond the point where it has proven to be effective means more people will die for no good reason. And tell me, what is morally acceptable about that?

No it doesn't. And no they aren't. It isn't proven. Something isn't a fact because you say so.

 queen_annes_revenge wrote:
No it doesn't. And no they aren't. It isn't proven. Something isn't a fact because you say so.

What is there to prove? He is saying that in the event that the technology meets the standards of the engineers (a standard that will be MUCH higher than a human driver) that to not adopt it is immortal. It's fine for you to say the tech will never get there but really. What gives you the idea that computers can't drive a car better than humans? Humans are pretty terrible drivers. just about every day there is an accident on my way to work. Those accidents could have been easily avoided if everyone was just paying attention. COMPUTERS ARE ALWAYS paying attention.

Are they? Is there any actual evidence to support this in a practical, real world scenario? Sure, on a small scale we know sensors work, and that individually systems can be made that are pretty good at doing certain things, but they are usually in specific closed environments, in niche roles, not interacting with the infinite different scenarios that may or may not present themselves in the public highway environment.

https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/812115

Well, here's some evidence that 94% of car accidents are caused by human error. If we can get machines to cause fewer accidents than that I think we're good.

And here's a very small sample from the self driving cars- where again, the humans are the problem.

http://fortune.com/2018/08/29/self-driving-car-accidents/

Conclusion- self driving cars are safer than humans. Please remember how low a bar that is.

Well the first statistic is irrelevant really. It doesn't prove the safety of autonomous systems.

And the bottom, well yeah of course they were, probably doing dumb stuff like the links I posted on Friday. And occurrences like that will increase massively the more you automate the system.

And like the other link I posted on Friday, Princeton studies predict that to obtain relevant safety data to allow a fair comparison with non autonomous vehicles will take 500 years.

 queen_annes_revenge wrote:
Are they? Is there any actual evidence to support this in a practical, real world scenario?

Auto Pilots on aircraft do pretty well. Sure they disengage for landing and take off, but my limited understanding is that a large amount of A to B is done via AP today for many commercial flights - though others here might have a more detailed understanding.

 Just Tony wrote:

Peregrine wrote:What happens with any field that becomes obsolete as technology advances? Seems like a good argument for socialism, but solutions to unemployment are rather off-topic from here.]And I pray that in 10 years time this will have been seen as a terrible idea and canned, with the research capabilities going to more productive endeavours. I can only hope.

Except that without employed people the companies will have nobody to purchase their robot made product, which will dry up resource supply AND that batch of other people's money that your socialist views depend on so desperately. No CEO is going to wake up and decide to give his money away for the greater good. If it came down to it, they'd take the entirety of their fortune and retire in Bora Bora, staring at caramel colored tiddies while all the socialists are forced to beat each other over the head for basic resources. We have historical examples of this happening. This isn't deluded fantasy, it's human condition.

Bad news for you, without employed people the companies will have nobody to puchase their robot-made product no matter what economic system you're under, leading to the same end result. Of course, if the CEO flees to Bora Bora his factories can be nationalized and started up manufacturing useful things for the people rather than let them all die pointlessly.

 queen_annes_revenge wrote:
Well the first statistic is irrelevant really. It doesn't prove the safety of autonomous systems.

And the bottom, well yeah of course they were, probably doing dumb stuff like the links I posted on Friday. And occurrences like that will increase massively the more you automate the system.

And like the other link I posted on Friday, Princeton studies predict that to obtain relevant safety data to allow a fair comparison with non autonomous vehicles will take 500 years.

Given the timeline of the Princeton data, that seems like an unrealistic standard. By that measurement, planes, cruise ships and cars are not proven safe. Indeed.... the transcontinental steamer won't be ok to analyze for another 300 years.

The 94% statistic definitely doesn't prove that automated vehicles are safe- but doesn't it indicate that humans are not? We get sloppy with routine tasks, we have bad days, we drive impaired, use cellphones, and routinely violate traffic laws intended to keep us safe.

I'd love to see city wide test set up, where human driven cars were not used for a year. Ideally, we'd do one with a similar population and in the same state as another city, then compare the two to see if our current generation self driving cars are safer than us. That should yield a reasonable standard to assess.

The great thing about computers is that they can be patched- every time one screws up, once its reported the problem can be fixed. They will become progressively safer over time. Humans.... won't.

I went back and looked at your example articles. All of them are about autopilot, or shared responsibility cars, which I'd say suffer from the worst of both worlds. Now you have a human, who is still nominally in charge, but a complacent one because someone else is supposed to be safely driving. So you've replaced the human with a computer, but then turned around and given override capability and ultimate responsibility to a human rendered inferior by their distracted, complacent state.

At the risk of the no true scotsman fallacy, I'd say that a car intended to be driven by humans, regardless of the sophistication of its digital enhancements, is not a self driving car. I think of self driving cars as vehicles in which I am a passenger, ideally where the driver has no ability to intervene or take control of the car under routine function.

 Vulcan wrote:

 Just Tony wrote:

Peregrine wrote:What happens with any field that becomes obsolete as technology advances? Seems like a good argument for socialism, but solutions to unemployment are rather off-topic from here.]And I pray that in 10 years time this will have been seen as a terrible idea and canned, with the research capabilities going to more productive endeavours. I can only hope.

Except that without employed people the companies will have nobody to purchase their robot made product, which will dry up resource supply AND that batch of other people's money that your socialist views depend on so desperately. No CEO is going to wake up and decide to give his money away for the greater good. If it came down to it, they'd take the entirety of their fortune and retire in Bora Bora, staring at caramel colored tiddies while all the socialists are forced to beat each other over the head for basic resources. We have historical examples of this happening. This isn't deluded fantasy, it's human condition.

Bad news for you, without employed people the companies will have nobody to puchase their robot-made product no matter what economic system you're under, leading to the same end result. Of course, if the CEO flees to Bora Bora his factories can be nationalized and started up manufacturing useful things for the people rather than let them all die pointlessly.

No, because the people who make their money off of products wouldn't let it get there in the first place. Remember, I work with automation all day every day, it's part of the deal in the machine shop. EVERY station that got automated still has the operator there.

 Peregrine wrote:

It doesn't make you a murderer, but it makes you responsible for those additional deaths.

Being responsible for someone's death is at best Manslaughter if not murder. So yeah, you are literally calling everybody who disagrees with you a murderer. That is both unclassy and a violation of rule 1.

 Just Tony wrote:

 Vulcan wrote:

 Just Tony wrote:

Peregrine wrote:What happens with any field that becomes obsolete as technology advances? Seems like a good argument for socialism, but solutions to unemployment are rather off-topic from here.]And I pray that in 10 years time this will have been seen as a terrible idea and canned, with the research capabilities going to more productive endeavours. I can only hope.

Except that without employed people the companies will have nobody to purchase their robot made product, which will dry up resource supply AND that batch of other people's money that your socialist views depend on so desperately. No CEO is going to wake up and decide to give his money away for the greater good. If it came down to it, they'd take the entirety of their fortune and retire in Bora Bora, staring at caramel colored tiddies while all the socialists are forced to beat each other over the head for basic resources. We have historical examples of this happening. This isn't deluded fantasy, it's human condition.

Bad news for you, without employed people the companies will have nobody to puchase their robot-made product no matter what economic system you're under, leading to the same end result. Of course, if the CEO flees to Bora Bora his factories can be nationalized and started up manufacturing useful things for the people rather than let them all die pointlessly.

No, because the people who make their money off of products wouldn't let it get there in the first place. Remember, I work with automation all day every day, it's part of the deal in the machine shop. EVERY station that got automated still has the operator there.

YOUR business may be that smart. I don't expect too many others to prioritize long-term thinking over this quarter's profits; thus far the vast majority does not.